Intuitive priors generator

Turn an intuitive guess — a single estimate plus a plausible range — into Beta(α, β) parameters you can drop straight into a Bayesian analysis.

Use this tool when:

You are studying a proportion, such as task completion rates or survey agreement.
You have an intuitive sense of the likely value: a best guess and a range of plausible outcomes.
For example, you might guess 80% task completion, and would be surprised below 70% or above 90%.

Why use a prior:

Ruling out unreasonable values sharpens inference — you may report 95% directional confidence instead of 90%.
A prior is mathematically equivalent to extra data, effectively increasing your sample size for free.
Uncomfortable with strong priors? Provide a very wide range to get a weak, nearly-flat prior.

Enter your assumptions

If I had to make a single estimate, the proportion is:

I would be surprised if the value is lower than:

...or higher than:

The shaded region shows the central 90% of the distribution — values within this band are where the true proportion is most likely to fall.